We have had several locklosses recently that can probably be attributed to a CSOFT P ring up. Starting on 6/8, I looked through all the locklosses and found 6 in total that had a CSOFT P ringup with the oscillation at 0.45 Hz. Some of these have already been noted in various alogs (shout out to Camilla for pointing them out here 70267).

CSOFT P locklosses:
6/8 2:45 UTC

6/8 6:53 UTC

6/10 6:53 UTC

6/10 19:50 UTC

6/11 21:45 UTC

6/12 12:12 UTC

The operating power for each of these locks seems to be about 430 kW in the X arm and 432 kW in the Y arm +/- 1 or 2 kW which is pretty standard, so it does not appear to be power related (at least lock to lock). The TMS servo signals also look stable and we are staying well centered on the TMS B QPDs and are not very far off center in the A QPDs.

Each of these locks have a different durations ranging from 2 to 9 hours, and there have been several other locks in this period where the locklosses are not attributed to ASC. Whatever the problem is, it does not appear to be consistent from lock to lock

We haven't had ASC problems for a while. These are some main differences from our known-to-be-stable configuration that could affect the ASC: ring heater change on EX, 1 W drop in input power, new input alignment from PR3 move.

One thing we could try is increasing the loop gain slightly and seeing if these locklosses persist. Right now the EPICs gain is set to 20. I have been able to increase this gain by up to 3 dB in the past without issues.